We have a knowledge base with near 1000 answers and a average of 15 alternate questions for each one of them. We are training the base both adding alternate questions as trying to indicate the correct answer using the portal.

We have struggling to improve the precision but without relevant success yet. An important problem is that base has a lot of similar answers between themselves, many times the difference among them is only an year (e.g. 2014 or 2018).

Some times when we use the GenerateAnswer API we need increase the parameter top a lot to receive the correct answer. For some questions we need assign the value 500 for top parameter, otherwise we get other answers with high scores, but still incorrect.

We thought that we could increase that results using the portal to train the knowledge base indicating the correct answer, but when we are trying it, we get just the first three top ranked answers, all incorrect.

*Note: We already migrated a test base to new approach (Language services/Question Answering) hoping that could be help us, but the problem stays, actually a bit worst.

How can we deal with this kind of situation? Can we change the top parameter value used by the portal? Are there another ways we can use to improve the rank?

Please provide us with the following information:

This issue is for a: (mark with an x)

- [X] bug report -> please search issues before submitting
- [ ] feature request
- [ ] documentation issue or request
- [ ] regression (a behavior that used to work and stopped in a new release)

Minimal steps to reproduce

To reproduce just follow steps in Batch test with tool at https://docs.microsoft.com/en-us/azure/cognitive-services/qnamaker/how-to/test-knowledge-base?tabs=v1#batch-test-with-tool
KB created with defined settings and from multi-turn sample.docx
** I have also got the same error with any other of my KBs, on the same QnA instance.

Any log messages given by the failure

For each line n sample.docx I got the following message:

Error processing line : 1, System.AggregateException: One or more errors occurred. (A task was canceled.)
---> System.Threading.Tasks.TaskCanceledException: A task was canceled.
--- End of inner exception stack trace ---
at System.Threading.Tasks.Task.ThrowIfExceptional(Boolean includeTaskCanceledExceptions)
at System.Threading.Tasks.Task`1.GetResultCore(Boolean waitCompletionNotification)
at batchtesting.Program.Main(String[] args) in C:\Work\gurvsing\cognitive-services-qnamaker-csharp\documentation-samples\batchtesting\Program.cs:line 18

Expected/desired behavior

results in out.tsv

OS and Version?

Windows 7, 8 or 10. Linux (which distribution). macOS (Yosemite? El Capitan? Sierra?)

Windows Server 2016 Datacenter

Versions

Mention any other details that might be useful

QnA Maker: Tier=Free F0
Search Service: Tier=Free, Replica 1, Partition= 1, Search Units= 1
App Service Plan: Tier=F1:Free, Apps/Slots= 1/0, OS=Windows
App Service: Tier=F1:Free

Thanks! We'll be in touch soon.

Hi,
How do i get all the questions in the qnaMaker via an API.
There are about 1000 questions in my QnaMaker Knowledge i need to fetch all the questions via an api and store them in a SqlServer Db.

Thank you.

Hello Team,

I am trying to replace a knowledge base(multi turn enabled). But in API call i am getting the Error saying invalid prompts given.
It will be great help if you will provide the sample Json for Replace a qnA qith multiturn enabled

This issue is for a: (mark with an x)

- [ ] bug report -> please search issues before submitting
- [ ] feature request
- [X ] documentation issue or request
- [ ] regression (a behavior that used to work and stopped in a new release)

Minimal steps to reproduce

Need to Add QnA Programmatically .. This returns UnAuthorized..
public async Task UpdateQueryQnABot(string newQuestion, string newAnswer)
{
string strResponse = "";
// Create the QnAKnowledgeBase that contains the new entry
QnAKnowledgeBase objQnAKnowledgeBase = new QnAKnowledgeBase();
QnaPair objQnaPair = new QnaPair();

        objQnaPair.question = newQuestion;
        objQnaPair.answer = newAnswer;
       // if (paramMode == Mode.Add)
       // {
            Add objAdd = new Add();
            objAdd.qnaPairs = new List<QnaPair>();
            objAdd.urls = new List<string>();
            objAdd.qnaPairs.Add(objQnaPair);
            objQnAKnowledgeBase.add = objAdd;
       // }

        using (System.Net.Http.HttpClient client =
            new System.Net.Http.HttpClient())
        {
            string RequestURI = String.Format("{0}{1}{2}{3}? ",
             @"https://westus.api.cognitive.microsoft.com/",
              @"qnamaker/v4.0/",
               @"knowledgebases/",
               knowledgeBaseId);
            //https://westus.api.cognitive.microsoft.com/qnamaker/v4.0/knowledgebases/
            //string RequestURI = qnaServiceHostName + "/qnamaker/knowledgebases/" + knowledgeBaseId ;
            //string questionJSON = @"{'question': '" + question + "'}";


            using (HttpRequestMessage request =
                new HttpRequestMessage(new HttpMethod("PATCH"), RequestURI))
            {
                request.Content = new StringContent(
                    JsonConvert.SerializeObject(objQnAKnowledgeBase),
                    System.Text.Encoding.UTF8, "application/json");
                request.Content.Headers.Add(
                    "Ocp-Apim-Subscription-Key", endpointKey);
                HttpResponseMessage response = await client.SendAsync(request);
                if (response.IsSuccessStatusCode)
                {
                    strResponse = $"Operation successfully completed.";
                }
                else
                {
                    string responseContent = 
                        await response.Content.ReadAsStringAsync();
                    strResponse = responseContent;
                }

Any log messages given by the failure

Expected/desired behavior

Please provide us with the following information:

This issue is for a: (mark with an x)

- [x ] bug report -> please search issues before submitting
- [ ] feature request
- [ ] documentation issue or request
- [ ] regression (a behavior that used to work and stopped in a new release)

Minimal steps to reproduce

Using the https://docs.microsoft.com/en-us/azure/cognitive-services/qnamaker/quickstarts/create-new-kb-csharp sample create simple question and answer where there is a simple

tag t the answer. Then exeecute the code.

Any log messages given by the failure

The simple

tag is removed when viewing the KB in QnAMaker.ai.

Expected/desired behavior

The new imported KB should have the simple

tag.

OS and Version?

Windows 7, 8 or 10. Linux (which distribution). macOS (Yosemite? El Capitan? Sierra?)

Versions

Windows 10

Mention any other details that might be useful

If I use the CSV import the tag is preserved or if I manually update the answer. I need to use the API since there is a log of content and doing it manually is not an option. The CSV option seems only to allow one question\answer which is also not an option since my KB has multiple questions.

---

Thanks! We'll be in touch soon.